CVE-2026-42888

Audiobookshelf is a self-hosted audiobook and podcast server. Prior to 2.36.0, the jwtAuthCheck function treats JWTs with the refresh token type as bearer access tokens on API and WebSocket resource endpoints such as /api/me instead of restricting them to /auth/refresh, allowing refresh tokens to authenticate as the associated users. This issue is fixed in version 2.36.0.

Vulnerability in audiobookshelf: authentication exemption check incorrectly matches unauthenticated GET routes, and CacheManager joins decoded values into cache file paths, potentially leading to unauthorized file access.

Prior to 2.33.0, a stored cross-site scripting (XSS) vulnerability exists in the Login Page due to improper sanitization of the authLoginCustomMessage field of the /api/auth-settings endpoint. An attacker with administrative privileges can inject arbitrary HTML/JavaScript that will be rendered on the login page for all users. This vulnerability is fixed in 2.33.0.

In Audiobookshelf prior to 2.32.2, the POST /api/backups/upload endpoint decompresses the details entry from an uploaded .audiobookshelf ZIP file entirely into memory using zip.entryData(), with no limit on decompressed size. The upload middleware also has no file size limit, so an admin user can upload a crafted ZIP containing a highly compressed details entry that, when decompressed, consumes hundreds of megabytes or gigabytes of memory, crashing the server process via out-of-memory. This vulnerability is fixed in 2.32.2.

In Audiobookshelf prior to 2.32.2, the POST /api/filesystem/pathexists endpoint uses String.startsWith() to validate that a resolved file path is within a library folder. The check fails for sibling directories sharing a common name prefix (e.g., /audiobooks vs /audiobooks-private), allowing authenticated users with upload permission to probe file existence outside their authorized library folder boundaries.

Audiobookshelf prior to 2.32.2 returns collections from all libraries via GET /api/collections and GET /api/collections/:id without verifying user access to each collection's library. An authenticated user with access to any library can enumerate and read collections (including full book metadata) from libraries they are explicitly restricted from.

In Audiobookshelf prior to 2.32.2, the GET /api/libraries/:id/download endpoint validates that the requesting user has access to the library specified in the URL path, but fetches downloadable items solely by attacker-provided IDs without constraining them to that library. An authenticated user with download permission and access to any one library can exfiltrate the full file contents of items belonging to any other library, including libraries they are explicitly denied access to.